I want to use ratio gain decision tree.but I dont know any thing about"confidence parameter".
I want to know that in which formula it used?
and what is the best value for this parameter?
another question is that does rapidminer support C5.0??
Solved! Go to Solution.
hello @e_serkani_e - welcome to the community. Having a decision tree with one node usually implies that your decision tree is not a useful model. It's saying that no matter what the inputs, it is always going to one node.
As for learning about confidence levels, there are many resources out there. I would highly recommend looking at a book on data mining to get a full explanation. The one I used many years ago, "Data Mining for the Masses" is still a free e-book and can be found here. Dr. North just republished it in a 2nd edition which is quite nice. There is also "Predictive Analytics and Data Mining" by Kotu & Deshpande which is excellent. Both books have numerous examples of RapidMiner processes and detailed explanations of all parameters e.g. confidence levels.